Research On The Key Technology Of Vision System In The Intelligent Three Coordinate Measuring Machine

Intelligent three coordinate measuring machine is the development trend of the newgeneration of the three coordinate measurement, it makes three coordinate measuringmachine more intelligent and automation in the application of the measurement anddetection. As the one of the key technology of the three coordinate measuring machine,the position and posture automatic identification system for parts has occupied a pivotalstatus in the developing process of the three coordinate measuring machine forintelligence, and the visual system is one of the most important method to realizeautomatic identification for the position and posture of the parts. At the same time，theresearch conents of the subject are the first part of the national natural science fund project(complex part measurement information generated and obtain mechanism and intelligentthree coordinates measuring key technology research), it is the foundation andprecondition of the follow-up work. In view of this, the key technology of the visualsystem in the position and posture to be automatic identification system in the subject isanalyzed and studied, and the development present situation of the intelligent threecoordinate measuring machine and its technology of position and posture to be automaticidentification for parts and visual measurement is been analyzed more comprehensive, themain content and realize the function of the position and posture automatic identificationsystem and the visual system in this paper is expounded, a identification method is putforward by using visual system for the placement position and three dimensionsinformation of the parts in the position and posture automatic identification system. Aim atthe identification content in the position and posture automatic identification system andthe visual system measurement method is put forward in the papers, with MATLAB forthe tool, the programming and validation make that the experiment process is moreintuitive and the feasibility of the proposed method is been verified.The main work accomplished listed in the paper include the following content:1) The composition and principle of the machine vision system and the position andposture automatic identification system for parts. The composition and principle of theparts position and posture automatic identification system is studied and the main task inthe process of identifying and the work by using visual system is analyzed. And the visualsystem was analyzed, of which the main research contents including visual systemcomposition, principle and mathematical model of the camera analysis. 2) Camera calibration. Camera is the key ingredients of the visual system, in order toanalyze and calculate image data gotten by camera more accurate, the camera must becalibrated first.3) Image processing. Parts image is picked up through the visual system, and then getthe data for recognition process through the image processing to carry on the followingoperation, the main image processing methods used include: image filtering, imagebinaryzation, edge extraction, images morphology analysis and so on.4) Parts placed posture recognition and the single camera binocular stereo visionmeasuring method for parts three dimensional information measurement. The parts placedway is recognized through the application of the BP neural network method, by theanalyzes of the stereo vision measuring principle, the single camera monocular stereovision measurement method in the three coordinate measuring machine is put forward forparts three dimensional coordinate information measurement.5) Use MATLAB for programming in order to realize the identification process andthe functions. With MATLAB for the tool, the parts the position and posture automaticidentification system operation program is designed and it realizes some functions such asthe real-time display image, image acquisition, image processing, camera parameterscalibration, placed posture identification, and three dimensional coordinate informationmeasurement etc.